fix(parsing.utils): hashtag-breaking '・' not escaped

Signed-off-by: Rongrong <[email protected]>

docs/CHANGELOG.md

@@ -11,6 +11,7 @@
 
 - **Canonical `DATABASE_URL` not recognized**: Since v2.9.0, `DATABASE_URL` is canonicalized before connecting to the corresponding database. However, a canonical URL pointing to a local path cannot be recognized when checking the validity of the scheme (database type). Both canonical (`scheme:/path/to/file.db`) and traditional (`scheme:///path/to/file.db`) forms of such URLs are recognized correctly now.
 - **Monitoring not deferred as per server-side cache when subscribing**: Since v2.7.0, monitoring tasks will be deferred when aggressive server-side caches (e.g., Cloudflare and RSSHub, which make it futile to check for updates before cache expiration) are detected. However, the first monitoring task for a newly subscribed feed was not being deferred. This has been fixed and the first monitoring task now waits for the server-side cache to expire.
+- **Minor bug fixes**
 
 ## v2.9.0: Telegraph-related revert, skip cert verification, and more
 

docs/CHANGELOG.zh.md

@@ -11,6 +11,7 @@
 
 - **无法识别规范的 `DATABASE_URL`**: 自 v2.9.0 起, 在连接到相应的数据库之前，`DATABASE_URL` 被规范化。然而，在检查 scheme (数据库类型) 的合法性时，无法识别指向本地路径的规范 URL。现在，此类 URL 的规范 (`scheme:/path/to/file.db`) 和传统 (`scheme:///path/to/file.db`) 形式都被正确识别。
 - **订阅时不会根据服务端缓存延迟监控**：自 v2.7.0 起，当检测到激进的服务器端缓存时，监控任务将被延迟（例如 Cloudflare 和 RSSHub，它们使得在缓存过期之前检查更新变得徒劳无功）。但是，当新订阅 feed 时，第一个监视任务不会被推迟。该问题已修复，第一个监控任务会等待服务端缓存过期。
+- **次要的 bug 修复**
 
 ## v2.9.0: 与 Telegraph 相关的 revert、跳过证书校验和更多
 

src/parsing/utils.py

@@ -94,9 +94,11 @@
     '\u2028'  # LINE SEPARATOR
     '\u2029'  # PARAGRAPH SEPARATOR
 )
-CHARACTERS_TO_ESCAPE_IN_HASHTAG: Final[str] = ''.join(
-    # all characters here will be replaced with '_'
-    sorted(set(SPACES + INVALID_CHARACTERS + string.punctuation + string.whitespace))
+INVALID_CHARACTERS_IN_HASHTAG: Final[str] = ''.join(
+    sorted(
+        # Known characters that break hashtags. Though '・' breaks hashtags, it is not the case of '·'.
+        set(chain(SPACES, INVALID_CHARACTERS, string.punctuation, string.whitespace, '・'))
+    )
 )
 
 escapeSpecialCharInReSet = partial(
@@ -155,7 +157,7 @@ def __merge_chars_into_ranged_set(sorted_chars: str) -> str:
     ' ',
 )
 escapeHashtag = partial(
-    re.compile(rf'[{__merge_chars_into_ranged_set(CHARACTERS_TO_ESCAPE_IN_HASHTAG)}]+').sub,
+    re.compile(rf'[{__merge_chars_into_ranged_set(INVALID_CHARACTERS_IN_HASHTAG)}]+').sub,
     '_',
 )
 isAbsoluteHttpLink = re.compile(r'^https?://').match